Here's a little supporting log from one of my tests.  This one creates 1000 
short lived processes
at the rate of 10 per second, then switches to only creating a few strings.

2017-05-03 11:28:07.160 Dtest[31914:891522] GC mark, 0 dead processes 
pending
2017-05-03 11:28:13.783 Dtest[31914:891522] GC mark, 0 dead processes 
pending
2017-05-03 11:28:14.027 Dtest[31914:891522] GC mark, 0 dead processes 
pending
2017-05-03 11:28:44.145 Dtest[31914:891522] GC mark, 293 dead processes 
pending
2017-05-03 11:29:14.209 Dtest[31914:891522] GC mark, 586 dead processes 
pending
2017-05-03 11:29:44.243 Dtest[31914:891522] GC mark, 879 dead processes 
pending
2017-05-03 11:30:14.320 Dtest[31914:891522] GC mark, 706 dead processes 
pending
2017-05-03 11:30:44.372 Dtest[31914:891522] GC mark, 413 dead processes 
pending
2017-05-03 11:31:14.424 Dtest[31914:891522] GC mark, 120 dead processes 
pending
2017-05-03 11:31:44.438 Dtest[31914:891522] GC mark, 0 dead processes 
pending

As you can see, the GCs occur at completely regular intervals.  The "706" 
line
corresponds to where new processes were no longer being spawned.

On Tuesday, May 2, 2017 at 11:13:46 PM UTC-7, Dave Dyer wrote:
>
> I've instrumented gcMark to show the size of the dead process queue.
> Gc seems to happen every 30 seconds like clockwork.
>
> On Tuesday, May 2, 2017 at 9:54:21 PM UTC-7, Shai Almog wrote:
>>
>> Spawns a couple isn't 100 per minute. A GC won't occur if RAM is in low 
>> use, thread garbage doesn't implicitly trigger a GC.
>>
>

-- 
You received this message because you are subscribed to the Google Groups 
"CodenameOne Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
Visit this group at https://groups.google.com/group/codenameone-discussions.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/codenameone-discussions/899a12f7-80ef-487e-99a4-80398f633bcd%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to